I just found a problem with the search. I've only tested 1.4.5 and 1.4.6 - both are affected.
When the search word is found in the html code of a document such as the word "font" it returns all documents that have font formatting html code. Viewing a document (with the highlight tag = font) it then starts showing the broken html tags with the word font like: font-weight: bold;">Error on display.
Not sure if this is fixed in newer versions of the script or not.
Problem searching with words found in the html code
Moderator: Thorsten
Hi,
this is a "bug", but how should I fix it? A fulltext search searches through all the words in the database field and so are HTML tags...
What do you think? A possibility could be filtering HTML tag names out of the search...
bye
Thorsten
this is a "bug", but how should I fix it? A fulltext search searches through all the words in the database field and so are HTML tags...
What do you think? A possibility could be filtering HTML tag names out of the search...
bye
Thorsten
phpMyFAQ Maintainer and Lead Developer
amazon.de Wishlist
amazon.de Wishlist
I'm not sure what you mean exactly by filtering.
Do you mean ignoring all text between < and > markers? I think that would give the most accurate results but that would probably affect the performance of the search quite a bit. Since < and > markers in the documents are converted to < and > when written to the database, FAQs on html code wouldn't be affected.
You could also mean excluding a list of words from the search like "FONT", "WIDTH", "TABLE", etc. Doing it this method would limit the types of documents that can be searched. If the FAQ contained documents about html coding you'd almost never see those documents in a search. If you decide to go with the word filter, I'd like to request that the list of words is available as an option in the configuration so that it can be modified easily to add or remove words if needed.
As to the best method - I'm not sure. How do other scripts using the WYSIWYG editor handle searches?
Do you mean ignoring all text between < and > markers? I think that would give the most accurate results but that would probably affect the performance of the search quite a bit. Since < and > markers in the documents are converted to < and > when written to the database, FAQs on html code wouldn't be affected.
You could also mean excluding a list of words from the search like "FONT", "WIDTH", "TABLE", etc. Doing it this method would limit the types of documents that can be searched. If the FAQ contained documents about html coding you'd almost never see those documents in a search. If you decide to go with the word filter, I'd like to request that the list of words is available as an option in the configuration so that it can be modified easily to add or remove words if needed.
As to the best method - I'm not sure. How do other scripts using the WYSIWYG editor handle searches?
Hi,
this is really a difficult one.
I don't know how other scripts handle HTML code and full text search...
bye
Thorsten
this is really a difficult one.
I don't know how other scripts handle HTML code and full text search...
bye
Thorsten
phpMyFAQ Maintainer and Lead Developer
amazon.de Wishlist
amazon.de Wishlist
there is also an other problem related to this:
When you seach a word that can be found in the html code (target of an image) It gets highlited, and the location of the image is fucked up.
problem could be solved by not searching text between < >
But how to do this is an other story
(I guess the search now uses a function of MySQL, maybe we have to look for a solution there?)
When you seach a word that can be found in the html code (target of an image) It gets highlited, and the location of the image is fucked up.
problem could be solved by not searching text between < >
But how to do this is an other story

Hi,
I know that and I'm looking for a solution...
bye
Thorsten
I know that and I'm looking for a solution...
bye
Thorsten
phpMyFAQ Maintainer and Lead Developer
amazon.de Wishlist
amazon.de Wishlist